KEF 104.2: still great after all these years!

[Sealed]

Kef 104.2 preliminary review



Description: D’appolito design mid/tweet/mid (MTM) with twin 8” woofers in a coupled chamber. 4th order advanced crossover.

Frequency response: 55hz-22khz +/- 2db. In room: about 30hz-22khz.

Efficiency: 92 db 1w/1 meter

Impedence: 4 ohms

Built from 1980’s… about 1987

Dimensions: 35.5” x 11 x 16.4 in 70.5 lbs each.

Original price 1986: about $2700 USD, extra for exotic finishes

Technology comments. This speaker is a timeless classic of accurate detail, dynamics and musical ability. The reason behind this is that KEF poured tons of money and engineering into it’s products from the 70’s and 80’s. The 103.x,104.x,105.x and 107.x were factory matched within 1db accuracy of factory reference. The box is miter type, excellently braced and very solid. The designers last product for KEF was the 105/4. After that product, KEF went on a downhill spiral. Even the new reference series are not as well engineered as the 10X.X era was. If these speakers were engineered and built like this now, you’d not be able to touch them for under $8k- $12k.

Tweeter: they are accurate and flat up to 22khz, matched within 0.5 db of each other. Model: T33 1" Impregnated fabric dome

Midranges: in MTM configuration they blend perfectly. Very holographic and detailed. The tweeter and mids are mounted on the front of the speaker in their own separate enclosure. Totally isolated from the woofer. 2 x 5" doped bextrene

Woofers: They are chamber loaded (meaning they are not ported, rather two push-pull drivers firing into a chamber that is open to the front. No coloration or noise is induced. The two woofers are coupled via force restoring rod. The bass is fast, tight, and very deep. 2 x 8" paper

Crossover: 4th order, built with tight tolerance components mounted on damped legs. This crossover is very complex, and includes baffle step, time and phase adjustments. May work with the Kef Universal Bass Equalizer (KUBE) to optimize bass below 100hz. Not required, but does enhance performance.

I found these in superior, ultra clean condition. I am the 2nd owner, and they are in a Rosewood finish. I paid $950.

Are these speakers perfect? Don’t ask that, no speaker is. Do these speakers deliver a huge, holographic, fast, rock-solid image with excellent impact, dynamics and low coloration? YES.

These speakers have been around, and are sought after and well loved. They do not take a back seat to any comparable sized speaker made now. They are on equal footing, if not less colored and more detailed than the Mighty Proac response 2.5.

What you can expect is a linear sounding speaker with great punch low enough that for 90% of music a sub isn’t necessary. These speakers like a good amount of solid state current. These speakers are revealing of very fine nuances, but are never sterile or analytical. They carry a tune with articulation and grace.

I know that Kef’s earlier work in the 70’s could be considered very colored, and polite. These are not like that at all. These could easily be used to master recordings, but they are not dry and lifeless like some monitor speakers can be.

I can without reservation say that these were made in Kef’s real heyday. They were privately and critically acclaimed, along with the 103.2,3,4 105.2,3,4 and 107. They are deserving of lofty praise indeed. With a competent front end, these will not sound was “romantic” or “slow” or “colored” as Kef has been accused of being. I would attribute those characterizations to the electronics and source of the day.

These speakers belong in an audio hall of fame, they are among the very best speakers made. They have no glaring faults, and the better the front end, the better they sound. Anyone looking for a great investment for around $1k, get a set of 104/2’s. long term owners will gladly corroborate how excellent these speakers were, and still are.

[oplancq]

[U][B]
These speakers belong in an audio hall of fame, they are among the very best speakers made. They have no glaring faults, and the better the front end, the better they sound. Anyone looking for a great investment for around $1k, get a set of 104/2’s. long term owners will gladly corroborate how excellent these speakers were, and still are.

Hello,
I also own some KEF 104/2. Bought those last year for 500 USD. They have a terrific soundstage and they are preety revealing of the rest of the system. Bad records are something to avoid !
Only fault is the medium 1KHz - 2.5KHz: this region is laid back; But they are terrific for the price I paid.


Olivier

[Sealed]

Hello,
I also own some KEF 104/2. Bought those last year for 500 USD. They have a terrific soundstage and they are preety revealing of the rest of the system. Bad records are something to avoid !
Only fault is the medium 1KHz - 2.5KHz: this region is laid back; But they are terrific for the price I paid.


Olivier

Mine are of 1985-85 vintage and I am experiancing no such problem. It could be your choice of electronics or cabling.

My friend has some 103/3's and those are very similar: smooth, flat and even.

KEF measurments (KEF and independant) measure the 104/2 absolutely ruler flat from 55hz to 22khz, +/- 2 db. Nothing is recessed at all.

[oplancq]

Mine are of 1985-85 vintage and I am experiancing no such problem. It could be your choice of electronics or cabling.

My friend has some 103/3's and those are very similar: smooth, flat and even.

KEF measurments (KEF and independant) measure the 104/2 absolutely ruler flat from 55hz to 22khz, +/- 2 db. Nothing is recessed at all.

I have tried several electronics, cables..always same result. In room measurement shows this region is slightly laid back. Mine are 1989 vintage.
Others 104/2 are showing same behaviour.

[oplancq]

By the way if you have documentation on those speakers I'd be quiet interested...

[Sealed]

By the way if you have documentation on those speakers I'd be quiet interested...

Suddenly you have many pairs of 104/2's in which you have measured the same thing? I find that dubious at best. Just suddenly you have access to many of them in a group test?

The kef literature is exact.

The 104.2 like to be *precisely* 50cm from back walls. Deviations of more than 1/4" are too much. They also must be 1 meter from side walls.

The speaker cables must not be more than .2ohms...period.

I have reason to believe that the 104/2 you have (1989) are NOT the same as the 1984-1987 version at all. Different drivers (like uni-q) and a different designer. The designer responsible for the whole batch was no longer at KEF after 1987.

It is entirely possible that your year 104/2 is totally different from mine.

It is a 100% certianty that mine do not exhibit any deviations in playback spectrum as you have stated.

[howyu@aol.com]

anybody know where I can get a set of tweeters for my KEF 104/2 speakers?
Thanks

[loganarch]

I am a newcomer to this forum, but I have been reading this thread with interest. I am not going to coment on the attitudes of the writers, but will rather comment on an experiment that I carried out today on two pairs of 104/2 which I came to acquire recently. One is older than the other, and I was comparing their sound and found out that they indeed sound very different.

The older pair has serial number 5028 and the newer one is 16121, what these mean in terms of years of production, I can't tell. The sound difference is significant though. The newer one has faster bass, much more detail, clarity and immediacy (I know these are not technical audiophile terms, but my subjective impression.) On the other hand, it also is too bright and hard on badly recorded material, female voice, symbals and some horns. I prefer the newer pair in well recorded symphonic classical music as the detail they provide can be absolutely breathtaking, while I prefer the older ones for voice and jazz, only because the harshness that I hear on the high end of the newer ones gets tiring with this sort of music. I suspect that this is a result of their extreme accuracy betraying recording flows and all kinds of other problems upstream.

I can confirm that the difference is in the crossover. Both boxes are identical in construction and drivers have the same SP numbers. Swapping the mid/high modules does not affect the sound at all (ie. the older box sounds just the same as it did before when mounted with the newer driver module). So the variable must be the crossover. I feel that the newer crossover is so transparent that many of the faults in the recording and the rest of the system shows through, and I know that the rest of my system is far from being at par with these speakers.

Having liked the detail and dynamics in the newer pair, I am planning to keep that one and see if I can soften the high end by better matching cables and amplifiers. So any advice would be very much appreciated. My current system includes a NAD 2200 amp, Rotel 975 CD player, NAD 1155 pre and 12 ga multistrand generic cables. Trying a NAD 2600 / NAD 1000 (from my home system - this is my office system) did not make an audible effect in the sound, changing the CD player to a Burr/Brown 1706 DAC did reduce some of the harshness. A 1970's Onkyo A50 integrated amp took away all of the sybillants and harshness, but at the same time the punch of the bass fizzled out and the midrange lost detail. Using 16 ga. generic wire had the same effect (I know this is not audiophile talk, but I was just experimenting to see what changes affect their sound in what way.)

Any advice on what type of amp and speaker wires to use?

My second pair, or parts thereof, will be for sale soon. They have one pair of good and one pair of fried woofers, (good inner surrounds) slightly scratched cabinets, good midranges and tweeters (one tweeter slightly dimpled but sound not affected by that) and one pair of good grills with pegs missing. I still have not decided whether to sell them in parts or as a whole, but if anyone here needs parts, let me know and if there is enough demand to justify parting them out, I will.

Yosi

[oplancq]

This is interesting...I was confirmed by KEF techs that there were a few different versions of
crossover used on the early 104/ 2 speakers. But the Xovers were fundamentally the same from a response, the difference being size and shape. Are your newer versions the biwire versions ? My 104/2 share the same behavior as your "new" ones: " The newer one has faster bass, much more detail, clarity and immediacy (I know these are not technical audiophile terms, but my subjective impression.) On the other hand, it also is too bright and hard on badly recorded material, female voice, symbals and some horns." Need to check what are the numbers. I don't know your system, but I believe KEFs like watts. I like my tube amp: VTL 75.


Olivier

[loganarch]

No, both of mine are the single wire versions.

Well, power it is not what's lacking. The NAD 2200 is rated 200W/ch into 4ohms, 440/ch into 2. and a headroom that doubles that. the NAD 2600 is even more powerful. So that is not the issue. These are solid state amps known for their "budget audiophile" performance, powerful, accurate and neutral, but nothing esoteric and possibly too much switching noise. I just purchased a Bedini 250/250 Class A 250W/ch amp and some Audioquest GBC speaker wire. The Bedini is known for tube-like smooth sound, while preserving some of the solid state characteristics such as fast bass and detail. I will let you know what I find out when I listen to the new setup.

I looked at the crossovers to the extent that can be seen through the woofers' port and they appear to be the same layout with capacitors of the same physical dimensions in the same locations. I am wondering if the difference in sound is due to deterioration of the crossovers. Maybe the newer one is closer to the way it is supposed to sound.

What is yours' serial number?

Yosi

[loganarch]

Olivier,

If you are talking to someone at KEF about this, it might be good to ask since both these speakers started their lives presumably with the same specificatioss for crossovers, which one represents deterioration, what kind of changes aging would cause in the crossovers.

Yosi

[oplancq]

Olivier,

If you are talking to someone at KEF about this, it might be good to ask since both these speakers started their lives presumably with the same specificatioss for crossovers, which one represents deterioration, what kind of changes aging would cause in the crossovers.

Yosi

Mine are 17607a/b..You can contact KEF through email. They are usually quiet responsive.
I agree that your NAD should have quiet enough power...Let me know the results with the Bedini.
Do you think that the voices are slightly laid back ?

Olivier